Deep Learning: Automated PD-L1 Tumor Cell Scoring of Resected NSCLC – Broadcast 2

Are you busy during this broadcast? See Broadcast #1 for an alternative time.

PD-L1 expression measured by immunohistochemistry helps identify non-small cell lung cancer (NSCLC) patients that may respond to anti-PD-1/PD-L1 immunotherapies. In this webinar, our featured speaker will present a novel deep learning solution for the automated scoring of PD-L1+ tumor cells (TC) in whole slide images of resected NSCLC.

A convolutional neural network (CNN) was used for the fine-grained classification of tissue regions into three classes: (1) regions of membrane-positive epithelial tumor cells [TC(+)], (2) regions of membrane-negative epithelial tumor cells [TC(-)], and (3) other regions that could wrongly influence scoring, i.e. macrophages, positive and negative lymphocytes, stroma and/or necrosis. The TC score was calculated as the ratio of the area of the classified TC(+) region to the sum of the areas of the classified TC(-) and TC(+) regions.

Two sets of ~225k training and ~30k testing patches (128×128 pixels) were created from manual partial annotations from (N=22) train and (N=12) test slides (Ventana-SP263). Training a fully convolutional network yielded maximum accuracy of 0.89 on test patches. The trained network was applied on (N=433) unseen confirmation slides and the TC score calculated for each slide based on the classified TC(+) and TC(-) regions. A non-linear gamma mapping to the manual TC scores by a trained pathologist was then estimated to maximize Overall Percent Agreement (OPA) at ≥25% criterion using two-fold cross-validation.

Evaluation against pathologist scoring on the confirmation slides yielded higher Overall (OPA), Negative (NPA) and Positive (PPA) Percent Agreement values at ≥25% criterion, higher Lin’s correlation and lower mean absolute error than a baseline approach relying on a heuristic detection of individual epithelium cells [ESMO-2017-103P]. Scoring by a second pathologist was available on a subset (N=170) of the confirmation slides. Average and standard deviation results on this subset confirm the above observations and suggest that the approach is getting close to inter-pathologist variability.

Join this webinar to learn how using deep learning to identify PD-L1 positive and negative tumor cell regions enables the automated scoring of PD-L1 TCs at the ≥25% expression level in resected NSCLC. These findings should be confirmed with additional tumor sets.

Speaker

Nicolas Brieu, PhD, Senior Research Scientist, Definiens

Nicolas has more than 10 years’ experience in the field of medical image analysis. His main area of research is in machine and deep learning for computer vision and data mining. He joined Definiens in 2011 and has since been working on the research and development of novel algorithms to automate the analysis of digital pathology images and the identification of immune-related prognostic factors.

Message Presenter